Maybe errors per man-hour stayed constant; errors per person went up because there are fewer employees.
Depends on how they're measuring. Unfortunately, measuring the right
thing is very hard. Management often can't measure the right thing, and end up basing their decisions on what they can
Example: A newly minted IT person followed all the recommendations from his classes: preventive maintenance, cleaned the machines, etc. Calls to tech support went down. Number of resolved issues per technician went down; management decided to reduce staff. Newly minted IT person gets axed.
What they should have been measuring is uptime and/or worker productivity, which went way up for a short while, but they had that handy little number....